The Brain Already Knew:
Why Teaching Kids to Design Decisions with AI Is Neurologically Overdue

A critical examination of the learn108 AI Kid approach through the lens of developmental neuroscience, cognitive science, and contemporary learning theory.

There is a peculiar gap at the centre of modern education. We teach children to read about decisions. We do not teach them to make them. We assign case studies, praise critical thinking as an abstraction, and then deliver children into a world of compounding uncertainty — algorithmic, social, economic — equipped with content but not with cognitive architecture. The learn108 AI Kid initiative, and its summer expression in the AI Kids Camp partnership with Pura Vida Basketball, proposes something quietly radical: that decision design — not coding, not AI literacy in the narrow technical sense, but the structured act of designing decisions with AI as a thinking partner — is the foundational skill of the coming generation. The neuroscience, it turns out, is firmly on their side.

The Prefrontal Cortex Problem

The most important piece of developmental neuroscience for this conversation is also the most underappreciated in curriculum design: the prefrontal cortex, the brain region responsible for executive function, deliberate decision-making, impulse regulation, and future-orientation, does not fully mature until the mid-twenties. This is not a marginal finding. It is one of the most replicated results in the entire field of developmental neurobiology, established across decades of research by scientists including Sarah-Jayne Blakemore at University College London and B.J. Casey at Weill Cornell Medicine.

What this means for children is not that they cannot make decisions — they can and do, constantly. It means their decision-making is disproportionately driven by the subcortical limbic system: fast, emotional, reward-seeking, present-biased. The prefrontal cortex is the slow system, the deliberate system, the system that asks “what happens next?” and “who else is affected?” It is precisely this system that decision design practice targets.

When the AI Kid framework teaches children to slow down, structure their thinking, and work through a decision architecture with an AI thinking partner, it is not merely delivering content. It is scaffolding the deliberate cognitive processes that the brain is still building. This is what developmental psychologists call “cognitive scaffolding” — providing external structure that gradually internalises as competence. Vygotsky called it the zone of proximal development: the sweet spot where a child cannot yet do something alone, but can do it with support. AI, deployed thoughtfully, is the most adaptive scaffolding tool ever created for that zone.

Metacognition: The Skill Nobody Teaches

Metacognition — thinking about thinking — is, per John Hattie’s landmark synthesis of over 800 meta-analyses of educational research, one of the highest-impact interventions available to educators, with effect sizes consistently above 0.6 (where 0.4 is considered a year’s worth of expected growth). And yet it remains largely incidental in most curricula, something students either develop on their own or don’t.

Decision design is metacognition made visible. When a child is asked to articulate what they know, what they assume, what they’re uncertain about, and what tradeoffs they’re accepting — which is precisely what working through a structured decision framework with an AI partner demands — they are practising metacognitive monitoring and control in real time. They are not just making a decision; they are watching themselves make a decision and adjusting the process as they go.

The AI component is not incidental to this process; it is the mirror. When a child articulates a decision problem to an AI and receives a structured response, they encounter their own thinking externalised. Cognitive science calls this “cognitive offloading” — using external systems to extend working memory and reduce cognitive load. But at the learn108 level, it is more than offloading; it is co-elaboration. The child’s thinking becomes visible to them in a way internal monologue never quite achieves. That visibility is what enables metacognitive growth.

Embodied Cognition and the Camp Model

Perhaps the most conceptually ambitious element of the learn108 approach is the AI Kids Camp format: combining basketball with AI and decision design. To a conventionally minded educator, this might look like an odd pairing. To a cognitive scientist, it is a sophisticated insight about how learning actually works.

Embodied cognition theory, advanced by researchers including Francisco Varela, Evan Thompson, and more recently Shaun Gallagher, holds that cognition is not a purely internal, brain-based process but is fundamentally shaped by the body’s interactions with its environment. Learning that is anchored in physical experience encodes differently — more durably, more accessibly, more emotionally resonant — than learning that is purely propositional.

Sport has always been implicitly understood as a laboratory for decision-making under pressure, but rarely has this been made pedagogically explicit. The AI Kids Camp format does exactly this: using the kinetic, emotionally heightened context of basketball to generate authentic decision problems, then using AI and the Decision Design framework to help children reflect on, decompose, and improve their decision processes. This loop — embodied experience, reflective structured analysis, re-entry into experience — mirrors what learning scientists describe as experiential learning cycles (Kolb) and what sports psychologists have long identified as the foundation of elite decision-making in athletes.

Crucially, it also addresses a persistent problem with classroom-based decision curricula: transfer failure. Skills learned in abstract, low-stakes contexts frequently fail to transfer to high-stakes real-world contexts. By anchoring decision design training in the physically and emotionally real context of sport, the camp model dramatically improves the conditions for transfer, because the emotional tagging and embodied encoding that sport naturally produces are precisely the conditions under which the brain retains and generalises learning.

Generation Alpha and the Ambient AI Environment

The children entering the AI Kid program are growing up in what researchers are beginning to call an “ambient AI environment” — a world in which AI-generated text, images, recommendations, and voices are not exceptional but pervasive. Generation Alpha, born from 2010 onward, will likely never know a world without AI-mediated information and decision surfaces.

The critical question this raises is not whether children will use AI — they will, and do, already. The question is whether they will use it as agents or as passengers. The learn108 positioning — “designing decisions with AI” rather than merely accepting AI outputs — is a direct intervention against the passenger mode. It cultivates what cognitive scientists call “epistemic agency”: the disposition and capacity to evaluate, question, and intentionally direct one’s own cognitive processes, including those mediated by external tools.

This is not a luxury skill. Research on algorithm aversion and algorithm appreciation (Dietvorst, Logg) consistently shows that untrained users of AI systems oscillate between over-reliance and under-reliance, rarely landing in the calibrated, critical engagement zone that actually produces good outcomes. Building that calibration in childhood — before habits of uncritical consumption calcify — is precisely the right developmental moment to intervene.

Multiple Intelligences and Learning Style Pluralism

Contemporary learning science has moved well beyond the original multiple intelligences framework (Howard Gardner, 1983), but the core insight — that children have meaningfully different cognitive profiles and that single-modality instruction systematically disadvantages many of them — has been vindicated by neuroscientific research into individual differences in working memory, processing speed, and attentional systems.

The learn108 AI Kid approach is, structurally, a multi-modal learning design. It engages linguistic-analytical intelligence through articulating and structuring decision problems. It engages spatial-visual intelligence through frameworks that map decisions architecturally. It engages bodily-kinaesthetic intelligence through the sports camp integration. It engages interpersonal intelligence through the team context of basketball. And it engages intrapersonal intelligence — arguably the most neglected of all in standard schooling — through the metacognitive reflection that is the engine of the whole enterprise.

Where most curricula still rely on a dominant auditory-linguistic mode, punctuated occasionally by visual aids, the AI Kid framework offers an approach that is genuinely pluralistic. More importantly, it offers one that is coherent: the different modes are not arbitrary variety but are all serving the same deep competency — decision literacy.

The Missing Curriculum: Decision Literacy as a Developmental Imperative

The central claim of the AI Kid initiative — that “designing decisions with AI may be the one thing kids will need most, yet it is the one thing many of us were never taught” — is worth taking seriously as a curriculum critique, not just as marketing language.

Comparative education research (PISA, TIMSS) consistently measures content knowledge and subject-specific reasoning. It does not measure decision literacy. National curricula around the world mandate science, mathematics, language, and increasingly digital literacy. None mandates decision design. And yet, as behavioural economists from Daniel Kahneman to Thaler and Sunstein have documented exhaustively, the quality of human decisions is the single greatest determinant of individual life outcomes — and the quality of human decision-making at scale is arguably the greatest determinant of societal outcomes as well.

The failure to teach decision design is not merely an oversight. It is a structural inheritance from an industrial model of schooling that valued compliance over deliberation, content delivery over cognitive architecture, and the reproduction of correct answers over the navigation of genuine uncertainty. The world AI Kids will inhabit is defined by the latter, not the former.

A Genuine Critical Note: Where the Execution Must Rise to the Ambition

An honest critical analysis must also identify where the promise of the approach creates obligations. The neurological and pedagogical case for decision design with AI is strong. But the strength of that case makes the quality of implementation more, not less, important.

The risk in any framework-driven learning program — particularly one that deploys AI as a pedagogical partner — is what cognitive scientists call “fluency illusion”: children and facilitators mistake the smooth, articulate outputs of AI-assisted thinking for genuine understanding, when in fact the thinking has been done by the tool rather than the learner. This is not an argument against AI in learning; it is an argument for careful facilitation design that consistently returns cognitive agency to the child. The framework must ask more of children over time, not less. The scaffold must be designed, from the beginning, to be removed.

Similarly, the embodied learning model of the camp is only as powerful as the quality of the reflective loop that connects the kinetic experience to the cognitive framework. Basketball without structured Decision Design reflection is just basketball — excellent, but not what is claimed. Decision Design without the embodied intensity of sport is just another classroom exercise. The synthesis is the value proposition, and the synthesis requires facilitation expertise that sits at the intersection of sport psychology, cognitive coaching, and AI literacy. That is a genuinely demanding profile to find in a single program delivery team.

These are not arguments against the approach. They are precisely the arguments for investing in the professional development of the educators and coaches who deliver it — which is, of course, where the Co-Cognition Champion framework becomes relevant again.

Conclusion: The Brain Was Always Ready for This

The neuroscience and learning science reviewed here do not merely permit the learn108 AI Kid approach. They demand something very much like it. A child’s developing prefrontal cortex is a decision-making system under construction. Metacognition is the highest-leverage educational intervention we know. Embodied learning produces the most durable transfer. Epistemic agency is the critical disposition for navigating an AI-saturated world. And decision literacy is the curriculum gap that virtually every other skill depends on but almost no institution directly addresses.

The AI Kid initiative is not filling a niche. It is, if the execution can match the conceptual ambition, addressing what may be the most consequential lacuna in modern childhood education. The brain, it turns out, has been waiting for this lesson. The only question is whether the institutions responsible for children’s development will recognise it in time to act.